Deciding on a Master's Thesis Direction for PhD Admissions in Canada

[Dragonfall]

I'm in Canada, where (usually) a master's degree, with a thesis, is required before being admitted in a PhD program. Now, I am told that a master's thesis does not need to be... substantial. However, you could say I'm ambitious (or stupid) and I want to get some original results. I'm in comp sci.

My question is this: should I rush out a mediocre thesis (ex., "An Implementation of So-and-So Algorithm in C++") then go for PhD, or try and do something substantial? I do plan on getting a PhD. Right now, I'm working on "lattice based cryptography", which apparently is very "hot".

 

[Choppy]

You know, I'm positive I've seen a PhD Comics about this exact topic.

What I would suggest is talking with prospective supervisors and picking a topic based on your interest rather than something you think you can get done quickly. You can't always judge how easy or fast a project is going to be (of course nor will you always enjoy what looks interesting, I suppose). But I would still choose based on your interests. Aiming for mediocre from the beginning doesn't give you much room to slack when things get tight.

 

[MATLABdude]

I'm also in Canada, and you can oftentimes (University and Department depending) "flip over" a Master's (a non-course based Master's, of course) into a Ph.D., provided the scope of the work is sufficient, and that you've done enough work to merit this step up, and that your Supervisor and Department agree with this assessment.

However, most who go on to a Ph.D. usually have a Master's intermediary, because it's the equivalent of passing go and collecting $200 (i.e. you have something in case you're unable to continue for whatever reason). I've known people who've done this, and gone and defended their Ph.D.s, and also people who've just wanted to get out, and dropped back down from a Doctoral program and back into a Master's (or even, in one case, a guy who dropped all the way into an M.Eng. program just to get the hell out with *something*)

Long story short, talk to your grad advisor for the options available to you, and your supervisor to see which of these options might suit you best based on your intended work.